Study of Co2 from 1700 to 2020 by country¶

  • Data Inspection

  • Data Cleaning

  • Visualization

  • Map Continent by Country

  • Dynamic render graph

    • Dynamic graph line co2/year by country

    • Dynamic graph cumultive co2 by country

    • Dynamic graph bar co2/year by country

    • Dynamic graph bar cumulative co2 by Continent

    • Dynamic graph lines cumulative co2 by Continent

  • Dynamic render function

    • function getPlot

    • function getFilter

    • function getPivot

    • function getRender

In [1]:
# Imports librairies

import pandas as pd
import matplotlib.pyplot as plt
import plotly.express as px
import math
import numpy as np
In [2]:
# import data from csv

df = pd.read_csv("co2_emission_by_countries.csv", encoding="latin")
df1 = df.copy()
df = pd.read_csv("continent_data.csv")
df2 = df.copy()

Data Inspection ¶

In [3]:
df1.describe()
Out[3]:
Year CO2 emission (Tons) Population(2022) Area
count 59620.000000 5.962000e+04 5.311600e+04 5.528400e+04
mean 1885.000000 1.034774e+09 3.992260e+07 6.522073e+05
std 78.231085 1.041652e+10 1.482365e+08 1.865483e+06
min 1750.000000 0.000000e+00 1.131200e+04 2.100000e+01
25% 1817.000000 0.000000e+00 1.770414e+06 1.770450e+04
50% 1885.000000 0.000000e+00 8.673095e+06 1.103815e+05
75% 1953.000000 8.715092e+06 2.862920e+07 4.925730e+05
max 2020.000000 4.170000e+11 1.425887e+09 1.709824e+07
In [4]:
df1.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 59620 entries, 0 to 59619
Data columns (total 9 columns):
 #   Column               Non-Null Count  Dtype  
---  ------               --------------  -----  
 0   Country              59620 non-null  object 
 1   Code                 57452 non-null  object 
 2   Calling Code         56097 non-null  object 
 3   Year                 59620 non-null  int64  
 4   CO2 emission (Tons)  59620 non-null  float64
 5   Population(2022)     53116 non-null  float64
 6   Area                 55284 non-null  float64
 7   % of World           55284 non-null  object 
 8   Density(km2)         53116 non-null  object 
dtypes: float64(3), int64(1), object(5)
memory usage: 4.1+ MB
In [5]:
df1.shape
Out[5]:
(59620, 9)
In [6]:
df1['Country'].value_counts().unique()
Out[6]:
array([271], dtype=int64)

Data cleaning ¶

In [7]:
df1 = df1.drop(columns=['Calling Code','Code'])
In [8]:
df1
Out[8]:
Country Year CO2 emission (Tons) Population(2022) Area % of World Density(km2)
0 Afghanistan 1750 0.0 41128771.0 652230.0 0.40% 63/km²
1 Afghanistan 1751 0.0 41128771.0 652230.0 0.40% 63/km²
2 Afghanistan 1752 0.0 41128771.0 652230.0 0.40% 63/km²
3 Afghanistan 1753 0.0 41128771.0 652230.0 0.40% 63/km²
4 Afghanistan 1754 0.0 41128771.0 652230.0 0.40% 63/km²
... ... ... ... ... ... ... ...
59615 Zimbabwe 2016 736467042.0 16320537.0 390757.0 0.30% 42/km²
59616 Zimbabwe 2017 746048675.0 16320537.0 390757.0 0.30% 42/km²
59617 Zimbabwe 2018 757903042.0 16320537.0 390757.0 0.30% 42/km²
59618 Zimbabwe 2019 768852126.0 16320537.0 390757.0 0.30% 42/km²
59619 Zimbabwe 2020 779383468.0 16320537.0 390757.0 0.30% 42/km²

59620 rows × 7 columns

Rename columns of select easier¶

In [9]:
df1 = df1.rename(columns={'CO2 emission (Tons)': 'coTons', 'Population(2022)' : 'population'})
In [10]:
df1
Out[10]:
Country Year coTons population Area % of World Density(km2)
0 Afghanistan 1750 0.0 41128771.0 652230.0 0.40% 63/km²
1 Afghanistan 1751 0.0 41128771.0 652230.0 0.40% 63/km²
2 Afghanistan 1752 0.0 41128771.0 652230.0 0.40% 63/km²
3 Afghanistan 1753 0.0 41128771.0 652230.0 0.40% 63/km²
4 Afghanistan 1754 0.0 41128771.0 652230.0 0.40% 63/km²
... ... ... ... ... ... ... ...
59615 Zimbabwe 2016 736467042.0 16320537.0 390757.0 0.30% 42/km²
59616 Zimbabwe 2017 746048675.0 16320537.0 390757.0 0.30% 42/km²
59617 Zimbabwe 2018 757903042.0 16320537.0 390757.0 0.30% 42/km²
59618 Zimbabwe 2019 768852126.0 16320537.0 390757.0 0.30% 42/km²
59619 Zimbabwe 2020 779383468.0 16320537.0 390757.0 0.30% 42/km²

59620 rows × 7 columns

Sorting the different columns to find correlations or anomalies¶

In [11]:
df1 = df1.sort_values('population', ascending = False)
In [ ]:

as we can see, we have a lot of NaN value let's see in 2020 how many Country contain NaN information in this data set¶

create filter for getting all data where year is 2020¶

In [12]:
dfSelect2020 = df1[df1['Year'] == 2020]
In [13]:
dfSelect2020
Out[13]:
Country Year coTons population Area % of World Density(km2)
10839 China 2020 2.360000e+11 1.425887e+09 9706961.0 6.30% 147/km²
24389 India 2020 5.441649e+10 1.417173e+09 3287590.0 2.00% 431/km²
57180 United States 2020 4.170000e+11 3.382899e+08 9372610.0 6.10% 36/km²
24660 Indonesia 2020 1.439996e+10 2.755013e+08 1904569.0 1.20% 145/km²
40649 Pakistan 2020 5.162289e+09 2.358249e+08 881912.0 0.50% 267/km²
... ... ... ... ... ... ... ...
45798 Saint Pierre and Miquelon 2020 3.676592e+06 NaN 242.0 0.00% NaN
48508 Sint Maarten (Dutch part) 2020 6.770353e+07 NaN NaN NaN NaN
53657 Timor 2020 6.543008e+06 NaN NaN NaN NaN
55554 Turks and Caicos Islands 2020 3.930547e+06 NaN 948.0 0.00% NaN
58806 Wallis and Futuna 2020 5.072470e+05 NaN NaN NaN NaN

220 rows × 7 columns

create filter for return all rows where one value have NaN on all columns¶

In [14]:
dfNull2020 = dfSelect2020[dfSelect2020.isnull().any(axis=1)]
In [15]:
dfNull2020['Country']
Out[15]:
23576                       Hong Kong
41191                       Palestine
39836                 North Macedonia
27912                          Kosovo
43901                         Reunion
9755                       Cape Verde
7858                           Brunei
1625                         Anguilla
1896                       Antarctica
2980                            Aruba
6232                          Bermuda
11110                Christmas Island
11923                           Congo
12194                    Cook Islands
12736                   Cote d'Ivoire
13549                         Curacao
14091                         Czechia
14362    Democratic Republic of Congo
17885                  Faeroe Islands
19240                French Polynesia
22492                   Guinea-Bissau
31164                           Macao
35771                      Montserrat
37939                   New Caledonia
39294                            Niue
44985                    Saint Helena
45798       Saint Pierre and Miquelon
48508       Sint Maarten (Dutch part)
53657                           Timor
55554        Turks and Caicos Islands
58806               Wallis and Futuna
Name: Country, dtype: object

how many are them ?¶

In [16]:
len(dfNull2020.index)
Out[16]:
31

pure curiosity, let's compare the most polluting to this group of countries¶

In [17]:
the31Union = dfNull2020['coTons'].sum()
theMostPoluating = dfSelect2020.nlargest(1, 'coTons')
In [18]:
x_value = [theMostPoluating['Country'].values[0], 'U34']
y_value = [theMostPoluating['coTons'].values[0], the31Union]

plt.bar( x_value, y_value)

plt.show()

I'm curious to find out the number of countries required to match the CO2 emissions of the United States¶

In [19]:
dfSorted2020 = dfSelect2020.sort_values('coTons', ascending=True)
dfSorted2020
Out[19]:
Country Year coTons population Area % of World Density(km2)
1896 Antarctica 2020 1.538880e+05 NaN NaN NaN NaN
43359 Puerto Rico 2020 2.088480e+05 3.252407e+06 8870.0 0.00% 367/km²
55825 Tuvalu 2020 2.789250e+05 1.131200e+04 26.0 0.00% 435/km²
39294 Niue 2020 2.901480e+05 NaN 260.0 0.00% NaN
44985 Saint Helena 2020 3.792950e+05 NaN 394.0 0.00% NaN
... ... ... ... ... ... ... ...
56909 United Kingdom 2020 7.816115e+10 6.750894e+07 242900.0 0.20% 278/km²
20324 Germany 2020 9.263562e+10 8.336984e+07 357114.0 0.20% 233/km²
44443 Russia 2020 1.150000e+11 1.447133e+08 17098242.0 11.00% 8/km²
10839 China 2020 2.360000e+11 1.425887e+09 9706961.0 6.30% 147/km²
57180 United States 2020 4.170000e+11 3.382899e+08 9372610.0 6.10% 36/km²

220 rows × 7 columns

In [20]:
i = 0
countriesSumsCo2 = 0
dfCountriesSums = pd.DataFrame(columns=df1.columns)
In [21]:
while theMostPoluating['coTons'].values[0] > countriesSumsCo2 :
    
    countriesSumsCo2 += dfSorted2020['coTons'].values[i] 

    row = pd.DataFrame([dfSorted2020.iloc[i]])
    dfCountriesSums = pd.concat([dfCountriesSums, row])

    
    i += 1
print(str(i) + " " + str(countriesSumsCo2))
    

207 418617496416.0
In [22]:
dfCountriesSums
Out[22]:
Country Year coTons population Area % of World Density(km2)
1896 Antarctica 2020 1.538880e+05 NaN NaN NaN NaN
43359 Puerto Rico 2020 2.088480e+05 3252407.0 8870.0 0.00% 367/km²
55825 Tuvalu 2020 2.789250e+05 11312.0 26.0 0.00% 435/km²
39294 Niue 2020 2.901480e+05 NaN 260.0 0.00% NaN
44985 Saint Helena 2020 3.792950e+05 NaN 394.0 0.00% NaN
... ... ... ... ... ... ... ...
7587 Brazil 2020 1.623851e+10 215313498.0 8515767.0 5.60% 25/km²
50134 South Korea 2020 1.833571e+10 51815810.0 100210.0 0.10% 517/km²
3251 Australia 2020 1.863575e+10 26177413.0 7692024.0 5.20% 3/km²
24931 Iran 2020 1.891444e+10 88550570.0 1648195.0 1.10% 54/km²
34416 Mexico 2020 2.007501e+10 127504125.0 1964375.0 1.30% 65/km²

207 rows × 7 columns

After observing this, i understand that CO2 emissions are cumulative from 1700 rather than being based annualy starting from 1700.¶

and when we see the united state at 417 billion tons co2Tons that confirm the data 'coTons' is cumulative and not based annualy¶

let's get 10 country who generated the most of c02 in 2020¶

In [23]:
df2020Big = dfSelect2020.nlargest(10, 'coTons')
In [24]:
df2020Big
Out[24]:
Country Year coTons population Area % of World Density(km2)
57180 United States 2020 4.170000e+11 3.382899e+08 9372610.0 6.10% 36/km²
10839 China 2020 2.360000e+11 1.425887e+09 9706961.0 6.30% 147/km²
44443 Russia 2020 1.150000e+11 1.447133e+08 17098242.0 11.00% 8/km²
20324 Germany 2020 9.263562e+10 8.336984e+07 357114.0 0.20% 233/km²
56909 United Kingdom 2020 7.816115e+10 6.750894e+07 242900.0 0.20% 278/km²
26557 Japan 2020 6.562517e+10 1.239517e+08 377930.0 0.20% 328/km²
24389 India 2020 5.441649e+10 1.417173e+09 3287590.0 2.00% 431/km²
18698 France 2020 3.872868e+10 6.462663e+07 551695.0 0.40% 117/km²
9484 Canada 2020 3.357674e+10 3.845433e+07 9984670.0 6.10% 4/km²
56367 Ukraine 2020 3.055916e+10 3.970174e+07 603500.0 0.40% 66/km²

Getting data from all year for the 10 selected¶

In [25]:
allYearCO2Big = df1[df1['Country'].isin(df2020Big['Country'])]
In [26]:
allYearCO2Big
Out[26]:
Country Year coTons population Area % of World Density(km2)
10839 China 2020 2.360000e+11 1.425887e+09 9706961.0 6.30% 147/km²
10754 China 1935 8.434728e+08 1.425887e+09 9706961.0 6.30% 147/km²
10740 China 1921 3.175662e+08 1.425887e+09 9706961.0 6.30% 147/km²
10741 China 1922 3.441998e+08 1.425887e+09 9706961.0 6.30% 147/km²
10742 China 1923 3.763514e+08 1.425887e+09 9706961.0 6.30% 147/km²
... ... ... ... ... ... ... ...
9309 Canada 1845 2.235040e+05 3.845433e+07 9984670.0 6.10% 4/km²
9310 Canada 1846 2.491520e+05 3.845433e+07 9984670.0 6.10% 4/km²
9311 Canada 1847 2.821280e+05 3.845433e+07 9984670.0 6.10% 4/km²
9483 Canada 2019 3.304092e+10 3.845433e+07 9984670.0 6.10% 4/km²
9349 Canada 1885 7.003736e+07 3.845433e+07 9984670.0 6.10% 4/km²

2710 rows × 7 columns

creating pivot_table for setting year as index and country as column for Emission though year on graph¶

In [27]:
pivot_table = allYearCO2Big.pivot_table(values='coTons', index='Year', columns='Country')

Emission over time from 1750 to now ¶

In [28]:
pivot_table.plot()
plt.ylabel('CO2 TONS')
plt.title('CO2 Emissions over time')
plt.show()

Let's see now how it was on the 19th century¶

In [29]:
dfCentury19 = allYearCO2Big[(allYearCO2Big['Year'] >= 1800) & (allYearCO2Big['Year'] <= 1900)]
In [30]:
pivot_table = dfCentury19.pivot_table(values='coTons', index='Year', columns='Country')
In [31]:
pivot_table.plot()
plt.ylabel('CO2 TONS')
plt.title('CO2 Emissions over time')
plt.show()

Let's see now how it was on the 20th century¶

In [32]:
dfCentury20 = allYearCO2Big[(allYearCO2Big['Year'] >= 1900) & (allYearCO2Big['Year'] <= 2000)]
In [33]:
pivot_table = dfCentury20.pivot_table(values='coTons', index='Year', columns='Country')
In [34]:
pivot_table.plot()
plt.ylabel('CO2 TONS')
plt.title('CO2 Emissions over time')
plt.show()
In [35]:
fig = px.line(pivot_table, x=pivot_table.index, y=pivot_table.columns[0:])
fig.show()

Let's see now how it was on the 21th century¶

In [36]:
dfCentury21 = allYearCO2Big[(allYearCO2Big['Year'] >= 2000) & (allYearCO2Big['Year'] <= 2020)]
In [37]:
pivot_table = dfCentury21.pivot_table(values='coTons', index='Year', columns='Country')
In [38]:
pivot_table
Out[38]:
Country Canada China France Germany India Japan Russia Ukraine United Kingdom United States
Year
2000 2.218015e+10 7.488403e+10 3.142863e+10 7.621667e+10 1.893972e+10 4.100939e+10 8.332305e+10 2.494736e+10 6.847098e+10 3.040000e+11
2001 2.273920e+10 7.839897e+10 3.183916e+10 7.713332e+10 1.993228e+10 4.225916e+10 8.483076e+10 2.525092e+10 6.904729e+10 3.100000e+11
2002 2.330387e+10 8.227156e+10 3.224517e+10 7.803329e+10 2.095531e+10 4.353828e+10 8.632645e+10 2.554660e+10 6.960736e+10 3.160000e+11
2003 2.388621e+10 8.681658e+10 3.265712e+10 7.893444e+10 2.201493e+10 4.482575e+10 8.785206e+10 2.585361e+10 7.017883e+10 3.220000e+11
2004 2.446697e+10 9.204033e+10 3.307027e+10 7.982153e+10 2.314040e+10 4.610861e+10 8.938275e+10 2.616398e+10 7.075173e+10 3.280000e+11
2005 2.504282e+10 9.791689e+10 3.348578e+10 8.068823e+10 2.432635e+10 4.739898e+10 9.093035e+10 2.647710e+10 7.132169e+10 3.340000e+11
2006 2.561423e+10 1.040000e+11 3.389154e+10 8.156655e+10 2.558610e+10 4.866634e+10 9.253690e+10 2.680977e+10 7.188926e+10 3.400000e+11
2007 2.620889e+10 1.110000e+11 3.428697e+10 8.241817e+10 2.694425e+10 4.996948e+10 9.414145e+10 2.714613e+10 7.244850e+10 3.470000e+11
2008 2.678826e+10 1.190000e+11 3.467496e+10 8.327310e+10 2.840706e+10 5.120179e+10 9.577837e+10 2.747166e+10 7.299300e+10 3.530000e+11
2009 2.733487e+10 1.270000e+11 3.504359e+10 8.406339e+10 3.001988e+10 5.236500e+10 9.730733e+10 2.774896e+10 7.348689e+10 3.580000e+11
2010 2.789368e+10 1.350000e+11 3.542017e+10 8.489634e+10 3.169777e+10 5.357982e+10 9.892044e+10 2.804303e+10 7.399853e+10 3.640000e+11
2011 2.846355e+10 1.450000e+11 3.577391e+10 8.570556e+10 3.347790e+10 5.484469e+10 1.010000e+11 2.835099e+10 7.446798e+10 3.690000e+11
2012 2.903281e+10 1.550000e+11 3.612984e+10 8.651954e+10 3.544148e+10 5.615070e+10 1.020000e+11 2.865496e+10 7.495550e+10 3.750000e+11
2013 2.960651e+10 1.650000e+11 3.648737e+10 8.735100e+10 3.747842e+10 5.746604e+10 1.040000e+11 2.895221e+10 7.543288e+10 3.800000e+11
2014 3.017707e+10 1.750000e+11 3.681326e+10 8.814359e+10 3.966428e+10 5.872977e+10 1.060000e+11 2.920972e+10 7.587142e+10 3.860000e+11
2015 3.075013e+10 1.840000e+11 3.714317e+10 8.893920e+10 4.193284e+10 5.995316e+10 1.070000e+11 2.943354e+10 7.629376e+10 3.910000e+11
2016 3.130970e+10 1.940000e+11 3.747675e+10 8.973988e+10 4.431507e+10 6.115688e+10 1.090000e+11 2.966755e+10 7.669310e+10 3.960000e+11
2017 3.187906e+10 2.040000e+11 3.781374e+10 9.052577e+10 4.674892e+10 6.234500e+10 1.100000e+11 2.989064e+10 7.708067e+10 4.010000e+11
2018 3.245853e+10 2.140000e+11 3.813612e+10 9.127988e+10 4.934873e+10 6.348846e+10 1.120000e+11 3.012267e+10 7.746256e+10 4.070000e+11
2019 3.304092e+10 2.250000e+11 3.845205e+10 9.199130e+10 5.197470e+10 6.459439e+10 1.140000e+11 3.034525e+10 7.783157e+10 4.120000e+11
2020 3.357674e+10 2.360000e+11 3.872868e+10 9.263562e+10 5.441649e+10 6.562517e+10 1.150000e+11 3.055916e+10 7.816115e+10 4.170000e+11
In [39]:
pivot_table.plot()
plt.ylabel('CO2 TONS')
plt.title('CO2 Emissions over time')
plt.show()

as we can see i have a problem on my X axis¶

In [40]:
pivot_table.plot(figsize=(15, 5), rot=30)
plt.xticks(ticks=pivot_table.index, labels=pivot_table.index.astype(int))
plt.ylabel('CO2 TONS')
plt.title('CO2 Emissions over time')
plt.show()

now we fixed the issue, let's go to the next step¶

well let's now see from this ten coutry who's generated most co2 by Area we gonna take 2020 data for the first step¶

In [41]:
df2020BigByArea = df2020Big
df2020BigByArea['coByArea'] = (df2020BigByArea['coTons'] / df2020BigByArea['Area'])
In [42]:
df2020BigByArea
Out[42]:
Country Year coTons population Area % of World Density(km2) coByArea
57180 United States 2020 4.170000e+11 3.382899e+08 9372610.0 6.10% 36/km² 44491.342326
10839 China 2020 2.360000e+11 1.425887e+09 9706961.0 6.30% 147/km² 24312.449592
44443 Russia 2020 1.150000e+11 1.447133e+08 17098242.0 11.00% 8/km² 6725.837662
20324 Germany 2020 9.263562e+10 8.336984e+07 357114.0 0.20% 233/km² 259400.681847
56909 United Kingdom 2020 7.816115e+10 6.750894e+07 242900.0 0.20% 278/km² 321783.226167
26557 Japan 2020 6.562517e+10 1.239517e+08 377930.0 0.20% 328/km² 173643.708930
24389 India 2020 5.441649e+10 1.417173e+09 3287590.0 2.00% 431/km² 16552.091045
18698 France 2020 3.872868e+10 6.462663e+07 551695.0 0.40% 117/km² 70199.441347
9484 Canada 2020 3.357674e+10 3.845433e+07 9984670.0 6.10% 4/km² 3362.829551
56367 Ukraine 2020 3.055916e+10 3.970174e+07 603500.0 0.40% 66/km² 50636.547377
In [43]:
df2020BigByArea.plot.bar(x='Country', y='coByArea', rot=30, figsize=(10,5))
plt.show()

let's get coByArea from 1700¶

In [44]:
allYearCoAreaBig = allYearCO2Big.copy()
allYearCoAreaBig['coByArea'] = (allYearCoAreaBig['coTons'] / allYearCoAreaBig['Area'])
In [45]:
allYearCoAreaBig
Out[45]:
Country Year coTons population Area % of World Density(km2) coByArea
10839 China 2020 2.360000e+11 1.425887e+09 9706961.0 6.30% 147/km² 24312.449592
10754 China 1935 8.434728e+08 1.425887e+09 9706961.0 6.30% 147/km² 86.893604
10740 China 1921 3.175662e+08 1.425887e+09 9706961.0 6.30% 147/km² 32.715307
10741 China 1922 3.441998e+08 1.425887e+09 9706961.0 6.30% 147/km² 35.459071
10742 China 1923 3.763514e+08 1.425887e+09 9706961.0 6.30% 147/km² 38.771292
... ... ... ... ... ... ... ... ...
9309 Canada 1845 2.235040e+05 3.845433e+07 9984670.0 6.10% 4/km² 0.022385
9310 Canada 1846 2.491520e+05 3.845433e+07 9984670.0 6.10% 4/km² 0.024953
9311 Canada 1847 2.821280e+05 3.845433e+07 9984670.0 6.10% 4/km² 0.028256
9483 Canada 2019 3.304092e+10 3.845433e+07 9984670.0 6.10% 4/km² 3309.164984
9349 Canada 1885 7.003736e+07 3.845433e+07 9984670.0 6.10% 4/km² 7.014489

2710 rows × 8 columns

create the pivot table for getting the graph¶

In [46]:
pivot_table = allYearCoAreaBig.pivot_table(values='coByArea', index='Year', columns='Country')
In [47]:
pivot_table.plot()
plt.ylabel('CO2 TONS')
plt.title('CO2 Emissions over time by Area')
plt.show()

i want to have a better vision with a dynamic graph, let's use plotly¶

In [48]:
fig = px.line(pivot_table, x=pivot_table.index, y=pivot_table.columns[0:])
In [49]:
fig.show()

manipulate data for transform Cumulated CO2 from 1750 to Year by Year CO2¶

In [50]:
df1['co2ForYear'] = math.nan
df1 = df1.sort_values(['Country','Year'], ascending=[True,True])
df1['co2ForYear'] = df1.groupby('Country')['coTons'].diff()
df1['co2ForYear'] = df1['co2ForYear'].fillna(df1['coTons'])
In [51]:
df1
Out[51]:
Country Year coTons population Area % of World Density(km2) co2ForYear
0 Afghanistan 1750 0.0 41128771.0 652230.0 0.40% 63/km² 0.0
1 Afghanistan 1751 0.0 41128771.0 652230.0 0.40% 63/km² 0.0
2 Afghanistan 1752 0.0 41128771.0 652230.0 0.40% 63/km² 0.0
3 Afghanistan 1753 0.0 41128771.0 652230.0 0.40% 63/km² 0.0
4 Afghanistan 1754 0.0 41128771.0 652230.0 0.40% 63/km² 0.0
... ... ... ... ... ... ... ... ...
59615 Zimbabwe 2016 736467042.0 16320537.0 390757.0 0.30% 42/km² 10737567.0
59616 Zimbabwe 2017 746048675.0 16320537.0 390757.0 0.30% 42/km² 9581633.0
59617 Zimbabwe 2018 757903042.0 16320537.0 390757.0 0.30% 42/km² 11854367.0
59618 Zimbabwe 2019 768852126.0 16320537.0 390757.0 0.30% 42/km² 10949084.0
59619 Zimbabwe 2020 779383468.0 16320537.0 390757.0 0.30% 42/km² 10531342.0

59620 rows × 8 columns

let's take the same country as before for see how it is for each year¶

In [52]:
mostCoByYear = df1[df1['Country'].isin(df2020Big['Country'].values)]
In [53]:
mostCoByYear
Out[53]:
Country Year coTons population Area % of World Density(km2) co2ForYear
9214 Canada 1750 0.000000e+00 38454327.0 9984670.0 6.10% 4/km² 0.000000e+00
9215 Canada 1751 0.000000e+00 38454327.0 9984670.0 6.10% 4/km² 0.000000e+00
9216 Canada 1752 0.000000e+00 38454327.0 9984670.0 6.10% 4/km² 0.000000e+00
9217 Canada 1753 0.000000e+00 38454327.0 9984670.0 6.10% 4/km² 0.000000e+00
9218 Canada 1754 0.000000e+00 38454327.0 9984670.0 6.10% 4/km² 0.000000e+00
... ... ... ... ... ... ... ... ...
57176 United States 2016 3.960000e+11 338289857.0 9372610.0 6.10% 36/km² 5.000000e+09
57177 United States 2017 4.010000e+11 338289857.0 9372610.0 6.10% 36/km² 5.000000e+09
57178 United States 2018 4.070000e+11 338289857.0 9372610.0 6.10% 36/km² 6.000000e+09
57179 United States 2019 4.120000e+11 338289857.0 9372610.0 6.10% 36/km² 5.000000e+09
57180 United States 2020 4.170000e+11 338289857.0 9372610.0 6.10% 36/km² 5.000000e+09

2710 rows × 8 columns

In [54]:
pivot_table = mostCoByYear.pivot_table(index = 'Year', values='co2ForYear',columns='Country' )
In [55]:
fig = px.line(pivot_table, x=pivot_table.index, y=pivot_table.columns[0:])
fig.show()
In [ ]:

As we can see, the measurements provided by the dataset are not precise enough. We will conclude with a comparison by continent and a prediction model for the coming years.¶

In [ ]:

Map Country with Continent ¶

In [56]:
df2test = df2[df2['Entity'].str.contains('united', case=False)]
df2test
Out[56]:
Entity Code Year Continent
263 United Arab Emirates ARE 2015 Asia
264 United Kingdom GBR 2015 Europe
265 United States USA 2015 North America
266 United States Minor Outlying Islands UMI 2015 Oceania
267 United States Virgin Islands VIR 2015 North America
In [57]:
df1['Continent'] = math.nan
In [58]:
df1.head(10)
Out[58]:
Country Year coTons population Area % of World Density(km2) co2ForYear Continent
0 Afghanistan 1750 0.0 41128771.0 652230.0 0.40% 63/km² 0.0 NaN
1 Afghanistan 1751 0.0 41128771.0 652230.0 0.40% 63/km² 0.0 NaN
2 Afghanistan 1752 0.0 41128771.0 652230.0 0.40% 63/km² 0.0 NaN
3 Afghanistan 1753 0.0 41128771.0 652230.0 0.40% 63/km² 0.0 NaN
4 Afghanistan 1754 0.0 41128771.0 652230.0 0.40% 63/km² 0.0 NaN
5 Afghanistan 1755 0.0 41128771.0 652230.0 0.40% 63/km² 0.0 NaN
6 Afghanistan 1756 0.0 41128771.0 652230.0 0.40% 63/km² 0.0 NaN
7 Afghanistan 1757 0.0 41128771.0 652230.0 0.40% 63/km² 0.0 NaN
8 Afghanistan 1758 0.0 41128771.0 652230.0 0.40% 63/km² 0.0 NaN
9 Afghanistan 1759 0.0 41128771.0 652230.0 0.40% 63/km² 0.0 NaN
In [59]:
df1[df1['Country'].isin(df2['Entity'])]
Out[59]:
Country Year coTons population Area % of World Density(km2) co2ForYear Continent
0 Afghanistan 1750 0.0 41128771.0 652230.0 0.40% 63/km² 0.0 NaN
1 Afghanistan 1751 0.0 41128771.0 652230.0 0.40% 63/km² 0.0 NaN
2 Afghanistan 1752 0.0 41128771.0 652230.0 0.40% 63/km² 0.0 NaN
3 Afghanistan 1753 0.0 41128771.0 652230.0 0.40% 63/km² 0.0 NaN
4 Afghanistan 1754 0.0 41128771.0 652230.0 0.40% 63/km² 0.0 NaN
... ... ... ... ... ... ... ... ... ...
59615 Zimbabwe 2016 736467042.0 16320537.0 390757.0 0.30% 42/km² 10737567.0 NaN
59616 Zimbabwe 2017 746048675.0 16320537.0 390757.0 0.30% 42/km² 9581633.0 NaN
59617 Zimbabwe 2018 757903042.0 16320537.0 390757.0 0.30% 42/km² 11854367.0 NaN
59618 Zimbabwe 2019 768852126.0 16320537.0 390757.0 0.30% 42/km² 10949084.0 NaN
59619 Zimbabwe 2020 779383468.0 16320537.0 390757.0 0.30% 42/km² 10531342.0 NaN

58807 rows × 9 columns

In [60]:
df1[~df1['Country'].isin(df2['Entity'])]['Country'].unique()
Out[60]:
array(['Faeroe Islands', 'Micronesia', 'Timor'], dtype=object)

as we can see we have 3 country missing¶

In [61]:
concat = pd.DataFrame({'Entity':['Faeroe Islands','Micronesia','Timor'], 'Code': math.nan, 'Year': math.nan, 'Continent':['Europe', 'Oceania', 'Asia']} )
In [62]:
df2 = pd.concat([df2, concat])
In [63]:
df2
Out[63]:
Entity Code Year Continent
0 Abkhazia OWID_ABK 2015.0 Asia
1 Afghanistan AFG 2015.0 Asia
2 Akrotiri and Dhekelia OWID_AKD 2015.0 Asia
3 Aland Islands ALA 2015.0 Europe
4 Albania ALB 2015.0 Europe
... ... ... ... ...
283 Zanzibar OWID_ZAN 2015.0 Africa
284 Zimbabwe ZWE 2015.0 Africa
0 Faeroe Islands NaN NaN Europe
1 Micronesia NaN NaN Oceania
2 Timor NaN NaN Asia

288 rows × 4 columns

In [64]:
df2.drop(['Code', 'Year'], axis=1, inplace=True)
In [65]:
df2.head()
Out[65]:
Entity Continent
0 Abkhazia Asia
1 Afghanistan Asia
2 Akrotiri and Dhekelia Asia
3 Aland Islands Europe
4 Albania Europe
In [66]:
df2.reset_index(drop=True, inplace=True)
In [67]:
df1['Country'].dtype
Out[67]:
dtype('O')

Mapping for each country he's Continent¶

In [68]:
df2_dic = df2.set_index('Entity')['Continent'].to_dict()
df1['Continent'] = df1['Country'].map(df2_dic)
In [69]:
df1.groupby('Country').first().head(10)
Out[69]:
Year coTons population Area % of World Density(km2) co2ForYear Continent
Country
Afghanistan 1750 0.0 41128771.0 652230.0 0.40% 63/km² 0.0 Asia
Albania 1750 0.0 2842321.0 28748.0 0.00% 99/km² 0.0 Europe
Algeria 1750 0.0 44903225.0 2381741.0 1.60% 19/km² 0.0 Africa
Andorra 1750 0.0 79824.0 468.0 0.00% 171/km² 0.0 Europe
Angola 1750 0.0 35588987.0 1246700.0 0.80% 29/km² 0.0 Africa
Anguilla 1750 0.0 NaN 91.0 0.00% None 0.0 North America
Antarctica 1750 0.0 NaN NaN None None 0.0 Antarctica
Antigua and Barbuda 1750 0.0 93763.0 442.0 0.00% 212/km² 0.0 North America
Argentina 1750 0.0 45510318.0 2780400.0 1.80% 16/km² 0.0 South America
Armenia 1750 0.0 2780469.0 29743.0 0.00% 93/km² 0.0 Asia

Dynamic render with function renderPlot ¶

getPlot Function ¶

In [70]:
def getPlot (df,category,type, plotType, mark):
        
    if plotType == 'bar':
        
        #sum value for getting value for continent 
        if category == 'Continent':
            df = df.groupby(['Year',category])[type].sum().reset_index()
        
        fig = px.bar(df, x='Year', y=type, color = category, barmode='group',  title=(type + ' By ' + category))
        fig.show()

    
    if plotType == 'line':
        fig = px.line(df, x=df.index, y=df.columns[0:], markers = mark,  title=(type + ' By ' + category))
        
        fig.show()

getPivot Function ¶

In [71]:
def getPivot(df, category, type):
    pivot_table = None

    #pivot table with sum of all '$type' have the same '$category' 
    if category == 'Continent':
        pivot_table = df.pivot_table(index='Year', columns=category, values=type, aggfunc='sum')

        
    else:
        pivot_table = df.pivot_table(index='Year', columns=category, values=type)
    
    return pivot_table

getFilter Function ¶

In [72]:
def getFilter(year,  category, range, list, df):

        # set on first letter for prevent error
        list = [item.capitalize() for item in list]

        #set a dataframe for a range of value
        if range == 'Yes' :
            df_filtered = df[df[category].isin(list) & df['Year'].between(year[0], year[1])]  

        #set a dataframe for a range of value take one or many value
        else :
            df_filtered = df[df[category].isin(list) & df['Year'].isin(year)]
            
        return df_filtered
In [73]:
"""

Year -> take a list of date to select one or many
Category -> If you need country data or ContinentData
Range -> take string if Yes : return year in a range, else : selected year only
df -> taking the dataframe 
list -> list of country
type -> take a list of continent or country
plotType ->'bar' or 'line'
marker if line is True
"""
Out[73]:
"\n\nYear -> take a list of date to select one or many\nCategory -> If you need country data or ContinentData\nRange -> take string if Yes : return year in a range, else : selected year only\ndf -> taking the dataframe \nlist -> list of country\ntype -> take a list of continent or country\nplotType ->'bar' or 'line'\nmarker if line is True\n"

render Function ¶

In [74]:
def renderPlot(year, category, range, list, df,type, plotType, mark):

    category = category.capitalize()
    
    filterData = getFilter(year, category, range, list, df)
      
    if plotType == 'line':
        BernardPivot = getPivot(filterData, category, type )
        return getPlot(BernardPivot,category,type, plotType, mark)

    else:
        return getPlot(filterData,category,type, plotType, mark)

Dynamic graph line co2/year by country ¶

In [75]:
renderPlot([1980,1995,1998,1999, 2000],'Country', 1, ['France', 'Belgium','Yemen'], df1,'co2ForYear', 'line', True)

Dynamic graph cumultive co2 by country ¶

In [76]:
renderPlot([1980,1995,1998, 2000],'Country', 1, ['France', 'Belgium','Yemen'], df1,'coTons', 'line', True)

Dynamic graph bar co2/year by country ¶

In [77]:
renderPlot([1980, 2000],'Country', 'Yes', ['France', 'belgium', 'canada'], df1,'co2ForYear', 'bar', True)

Dynamic graph bar cumulative co2 by Continent ¶

In [78]:
renderPlot([1995, 2002, 2000],'Continent', 'Yes', ['europe', 'asia'], df1,'coTons', 'bar', True)

Dynamic graph lines cumulative co2 by Continent ¶

In [79]:
renderPlot([1995, 2002],'Continent', 'Yes', ['europe', 'asia'], df1,'coTons', 'line', True)
In [ ]: